Send to

Choose Destination
See comment in PubMed Commons below
Fertil Steril. 2002 Sep;78(3):614-8.

Reliability of clinicians versus radiologists for detecting abnormalities on hysterosalpingogram films.

Author information

  • 1Department of Obstetrics and Gynecology, Pennsylvania Hospital, Philadelphia 19104, USA.



To evaluate the consistency of the identification of abnormal findings on hysterosalpingogram (HSG) and compare the reliability of clinicians to that of radiologists.


Evaluation of reliability of diagnostic test.


Women undergoing evaluation for infertility.INTEVENTION(S): Retrospective review of 50 HSG films by three reproductive endocrinologists and three radiologists. Each film was reread 30 days later in a blinded fashion.


The consistency of each individual reader, the reliability of detecting specific abnormalities, and the consistency of clinicians compared with radiologists was evaluated with a kappa (K) statistic and interclass correlation coefficient (ICC).


Average intrareader reliability was high for the detection of normal uterus, normal tubes, and tubal obstruction and low for the detection of hydrosalpinx, uterine adhesions, and pelvic adhesions. Inter-reader reliability was high in the detection of normal uterine contour, normal tubal patency, and uterine filling defect and lower for the detection of a hydrosalpinx. The reliability of detecting pelvic adhesion or salpingitis isthmica nodosa was poor.


Intrareader reliability was generally good, especially for the detection of normal findings. Agreement among different readers is lower in detecting rare outcomes such as hydrosalpinx and pelvic adhesion and salpingitis isthmica nodosa. Clinicians more reliably diagnose hydrosalpinx and tubal obstruction, while radiologists more reliably detect the more subtle findings of salpingitis isthmica nodosa or uterine adhesions.

[PubMed - indexed for MEDLINE]
PubMed Commons home

PubMed Commons

How to join PubMed Commons

    Supplemental Content

    Full text links

    Icon for Elsevier Science
    Loading ...
    Write to the Help Desk